Thrombopoietin (TPO) is a recently isolated ligand of mpl (Bartley et al. (1994) Cell 77:1117; Kaushansky et al. (1994) Nature 369:568; Lok et al. (1994) Nature 269:565; Kuter et al. (1994) Proc. Natl. Acad. Sci. USA 91:11104-11108; Kuter & Rosenberg (1994) Blood 84:1464; Wendling et al. (1994) Nature 369:571), first identified as the proto-oncogene transduced by the murine myeloproliferative leukemia (MPL) virus (Wendling et al. (1989) Blood 73:1161-1167; Souyri et al. (1990) Cell 63:1137-1147; Vigon et al. (1992) Proc. Natl. Acad. Sci. USA 89:5640-5644; Skoda et al. (1993) EMBO J. 12:2645-2653; Methia et al. (1993) Blood 82:1395-1401). TPO has been shown to independently stimulate megakaryocyte (MK) progenitor division and MK maturation in vivo and in vitro (Bartley et al. (1994) supra; Kuter et al. 1994) supra; Kuter & Rosenberg (1994) supa; Wendling et al. (1994) supra; de Sauvage et al. (1994) Nature 369:533; Broudy et al. (1995) Blood 85:1719-1726; Lok & Foster (1994) Stem Cells 12:586-598; Zeigler et al. (1994) Blood 84:4045). In vivo administration of TPO to thrombocytopenic rodents was found to significantly boost the platelet count as well as increase the number and ploidy of maturing MKs in the bone marrow (Lok et al. (1994) supra; Lok et al. (1994) supra; Wendling et al. (1994) supra; de Sauvage et al. (1994) supra). Absence of a c-mpl (TPO receptor) gene in mice was reported to result in thrombocytopenia (Guerney et al. (1994) Science 265:1445). More recently, it has been demonstrated that MKs can be primed to produce functional platelets in culture after exposure to TPO (Choi et al. (1995) Blood 85:402).
Hematopoietic stem cells are rare cells that have been identified in fetal bone marrow, umbilical cord blood, adult bone marrow, and peripheral blood, which are capable of differentiating into each of the myeloerythroid (red blood cells, granulocytes, monocytes), megakaryocyte (platelets) and lymphoid (T-cells, B-cells, and natural killer cells) lineages. In addition, these cells are long-lived, and are capable of producing additional stem cells, a process termed self-renewal. Stem cells initially undergo commitment to lineage restricted progenitor cells, which can be assayed by their ability to form colonies in semisolid media. Progenitor cells are restricted in their ability to undergo multi-lineage differentiation and have lost their ability to self-renew. Progenitor cells eventually differentiate and mature into each of the functional elements of the blood. This maturation process is thought to be modulated by a complex network of regulatory factors including erythropoietin (EPO), granulocyte colony stimulating factor (G-CSF), granulocyte-macrophage colony stimulating factor (GM-CSF), thrombopoietin (TPO), steel factor (Stl), Flk-2 ligand and interleukins (IL) 1-15.
Recently, in vitro assays have been developed to identify human hematopoietic stem cells having self-renewal and multi-lineage differentiative capacity. One assay is the cobblestone area-forming cell (CAFC) assay, based on freshly isolated stromal cells or established stromal cell lines. In the mouse system, late-appearing cobblestone area formation on fresh marrow-derived stroma (Ploemacher et al. (1991) Blood 78:2527) or on a cloned stromal cell line (Neben et al. (1993) Exp. Hematol. 21:438) has been shown to correlate with in vivo hematopoietic repopulating ability. Correlation of CAFC and long term culture-initiating cell (LTCIC) frequencies using the mouse stromal cell line SyS1 has been demonstrated (Reading et al. (1994) Exp. Hematol. 22:406). In addition, the in vivo severe combined immunodeficiency (SCID)-hu bone assay has been used to measure long term engraftment of candidate stem cell populations (Kyoizumi et al. (1992) Blood 79:1704; Baum et al. (1992) Proc. Natl. Acad. Sci. USA 89:2804; Chen et al. (1993) Blood 82 (Suppl. 1):180a). Both the in vitro CAFC assay and the SCID-hu bone model permit analysis of B-cell and myeloid cell generation from candidate pluripotent hematopoietic stem cells (PHSC).
It is becoming increasingly apparent that distinct subpopulations of stem cells may be responsible for different phases of engraftment post transplantation. As early as 1964, differences in the ability of murine spleen colony forming units (CFU-S) to generate secondary CFU-S were defined (Ploemacher & Brons (1994) Exp. Hematol. 17:263-266). Although evidence now indicates that most CFU-S are not involved in repopulating lethally irradiated hosts (Jones et al. (1990) Nature 347:188-189; Jones et al. (1989) Blood 73:397-401), heterogeneity in transplantation potential appears to exist even within subpopulations of radioprotective cells. This has been demonstrated with serial bone marrow transplantations. The long-term repopulating ability of the grafts are lost with serial transfers, while a cell population survives which contributes to short-term reconstitution (Jones et al. (1989) supra). Further support for the concept that both short-term and long-term reconstituting stem cell populations exist have been derived from studies in which isoenzyme analysis and retroviral gene marking of hematopoietic cells have been utilized to track the fate of stem cells. A mathematical analysis of correlations and variances of donor reconstitution with isoenzyme variants in lethally irradiated mice indicates that a large number of multi-lineage clones are active immediately after reconstitution but rapidly decline, with the majority being inactive 12 weeks post-transplantation (Harrison & Zhong (1992) Proc. Natl. Acad. Sci. USA 89:10134-10138; Harrison et al. (1993) Exp. Hematol. 21:206-219). These observations indicate the existence of a population of cells with multi-lineage short-term engrafting potential in donor murine bone marrow. Similar observations have been made in a. large animal transplantation model, where isoenzyme differences have indicated the contribution of multiple clones to short-term engraftment followed by sustained contributions by relatively few stem cell clones (Abkowwitz et al. (1990) Proc. Natl. Acad. Sci. USA 87:9062-9066). These findings have been confirmed by an eloquent analysis of clonal development after transplantation with retrovirally marked stem cells (Jordan et al. (1990) Cell 61:953; Capel et al. (1990) Blood 75:2267).
Theoretically, subsets of cells with differing proliferative potentials may also differ with regards to physical characteristics, and therefore may be isolated and functionally defined. Bone marrow cells responsible for reconstitution following lethal irradiation can be isolated using centrifugation techniques exploiting cell size and density fractionation, or fluorescence-activated cell sorting (FACS) based on uptake of fluorescent vital dyes, lectin binding, or cell surface antigen expression (Sprangrude (1989) Immunol. Today 10:344-350; Visser & Van Bekkum (1990) Exp. Hematol. 18:248-256). FACS isolated murine cells that are responsible for engraftment lack lineage markers for B-cells, T-cells, myelomonocytes and erythrocytes and are termed lineage negative (Lin−) (Spangrude et al. (1988) Science 241:58-62). The Lin− fraction of murine bone marrow can be further subdivided based on low levels of Thy-l and expression of the Sca-1 antigen (Visser & Van Bekkum (1990) supra; Spangrude et al. (1988) supra; Szilvassay et al. (1989) Blood 74:930-939; Szilvassay & Cory (1993) Blood 81:2310-2320; Spangrude & Scollay (1990) Exp. Hematol. 18:9920-926; Jurecic et al. (1993) Blood 82:2673-2683). Murine cells that are Thy-1.1loLin−Sca-1+ are 1000-fold enriched in radioprotective ability, and contain all of the radioprotective cells found in the bone marrow of syngeneic C57BL/thy-1.1 mice (Uchida & Weissman (1992) J. Exp. Med. 175:175-184). As few as 100 cells that are Thy1.1loLin−Sca-1+ can radioprotect at least 95% of lethally irradiated recipients with long term donor derived reconstitution. At this cell dose, Thy1.1loLin−Sca-1+ cells give rise to donor peripheral blood white blood cells by 10 days post transplant, and to platelets within 14 days of transplant (Uchida et al. (1994) Blood 83:3758-3779). These studies suggest this population contains cells with both short-term and long-term engrafting potential.
Thy1.1loLin−Sca-130  cells can be further divided by heterogeneity of cell cycle status and uptake of a fluorescent dye, rhodamine 123 (Spangrude & Johnson (1990) Proc. Natl. Acad. Sci. USA 87:7433-7437; Li & Johnson (1992) J. Exp. Med. 175:1443-1447; Fleming et al. (1993) J. Cell Biol. 122:897-902; Wolf et al. (1993) Exp. Hematol. 21:614-622). Rhodamine 123 is a dye that identifies active mitochondria, and its efflux from the cell is handled by the multidrug resistance gene product, P-glycoprotein. Those cells that retain small quantities of rhodamine 123 are termed rhodamine dull (or low) and have been shown to possess marrow repopulating ability (MRA) (Li & Johnson (1992) supra).
The cells responsible for reconstituting hematopoiesis in humans receiving a bone marrow transplant reside in a subset of cells expressing the CD34 antigen (CD34+) (Berenson et al. (1991) Blood 77:1717-1722). This fraction of cells can be further subdivided based on multiple antigen characteristics (Lansdorp et al. (1990) J. Exp. Med. 172:363-366; Verfaille et al. (1990) J. Exp. Med. 172:509-520; Briddell et al. (1992) Blood 79:3159-3167) including the lack of lineage specific markers (Lin−) and expression of the Thy antigen (Thy-1+) (Baum et al. (1992) Proc. Natl. Acad. Sci. USA 89:2804-2808; Craig et al. (1993) J. Exp. Med. 177:1331-1342; Murray et al. (1990) Blood Cells 20:364-370; Murray et al. (1995) Blood 85:468). In vivo assays using adult bone marrow and mobilized peripheral blood cells that are CD34+Thy-1+Lin− have shown that this population contains cells capable of contributing to all hematopoietic lineages (Chen et al. (1994) Blood 84:2497-2505; Galy et al. (1994) 84:104-110).
Mpl expression has been detected by polymerase chain reaction (PCR) in human hematopoietic cells throughout the MK lineage, as well as in primitive CD34+CD38lo/− cells (Debili et al. (1995) Blood 85:391-401). However, little is known about the actions of TPO on primitive cells prior to commitment to the MK lineage. One report has shown that primitive mouse HSC of the phenotype Sca-1+AA4+ express mpl, and exposure of these cells to TPO in vitro resulted in MK differentiation (Zeigler et al. (1994) supra).